Normalization

This article explains normalization in databases, its steps, and importance. It provides an example of normalization and discusses the issues that arise without normalization. The article also covers denormalization, business rules, and their impact on normalization and denormalization.

Running head: NORMALIZATION 1
Normalization
Student_First_Name Student_Last_Name
Institution

NORMALIZATION
Normalization
Fong defines normalization as the process of arranging data in a database to evade data
insertion, update anomaly, deletion anomaly and redundancy (Fong, 2015). Normalization
steps are outlined below:
1. Choose data origin and convert it to un-normalized entity.
2. Transform the un-normalized data to 1NF by the rule of removing any repeated
attributes to a new entity table.
3. Transform the data in 1NF to 2NF by the rule of removing every non-key fields that
are only dependent on part of table to new entity table.
4. Transform the data in 2NF into 3NF by the rule of removing to a new entity any non-
primary key.
Example normalization process
Stu_ID Stu_
Name
Tuitio
n Fees
Paid
($)
Course
Title
Class A Class B Class C
1 Susan
John
400 Medicine Botany 2
2 Scott
Smith
200 Economics Economics
1
Botany 1
3 Matt
Longson
850 Dentistry
4 Mary
Griffin
500 Computer
Engineering
Botany 1 Marketing
Intro
Programming 2
This isn’t a normalized entity, few issues exist within the table as outlined below.
Insert anomaly, which occurs when we insert a record into our table with no good
knowledge of data we are required to know i.e. we are required to add new student info but
we do not know their course title. New row will row will look like this

NORMALIZATION
5 Jared
Mazq
0 ?
Update anomaly, which occurs when we update the table, and some of the data is updated
leaving other data. For instance, class Botany was altered to Intro Botany. We will be
required to query every field that has this class column and alter it to match it (Aa, 2018).
First normal form
To convert our table into 1NF, we need to ask ourselves the following questions, 1) are
columns combined together make unique record at each single moment, 2) which column is
used to primarily identify the record? Looking at the first question, Ans. is NO. There would
be same combination of data that will signify a different record. Looking at the second
question, Ans. is NO, as there is a likelihood that more than one students may have same
name. We address this by creating new PK. For instance
Student (Stu_ID, Stu_ Name, Tuition _Fees_Paid ($), Course_Title, Class_A,Class_B_Class
C).
This table is now in 1NF.
Second normal form
An entity is said to be in 2NF when it fulfils 1NF requirements and every non-primary key
does not functionally depend on PK. Looking at our table,
Student (Stu_ID, Stu_ Name, Tuition _Fees_Paid ($), Course_Title, Class-
_A,Class_B_Class C), are all columns dependent on PK. Stu_ Name and Tuition _Fees_Paid
($) are dependent on Stu_ID PK, but Course_Title,Class_A,Class_B and Class C are not. We
have to fix this by coming up with student, course and class tables as below.

NORMALIZATION
Student (Stu_ID, Stu_ Name, Tuition _Fees_Paid ($))
Course (Course_ID, Course_Title)
Class (Class_ID, Class_Name)
Third normal form
A DataBase is said to be in 3NF if it fulfils 2NF requirements and does not have transitive
functional dependency. By transitive dependency we mean, field 1 determines field 2 which
determines filed 3. This has to be removed. When we look at our tables, none has this feature.
Therefore our tables are in 3NF.
Denormalization is required to reduce the scaling issues, address performance issues and
complexity of queries.
Business rules have an influence on both normalization and denormalization processes.
Normalization and denormalization have always to adapt to the dynamic business rules
avoiding change of data minus restructuring of the whole system. This implies that any
change of business rules, will result to changed normalization and denormalization (Hillard,
2010).

NORMALIZATION
References
Aa, H. v. (2018). Comparing and Aligning Process Representations: Foundations and
Technical Solutions. New York City: Springer.
Fong, J. S. (2015). Information Systems Reengineering, Integration and Normalization. New
York City: Springer.
Hillard, R. (2010). Information-Driven Business: How to Manage Data and Information for
Maximum Advantage. Hoboken: John Wiley & Sons.